Академический Документы
Профессиональный Документы
Культура Документы
ORIGINAL ARTICLE
Department of Computer Sc. & Technology, College of Information Sc. & Engineering, Ocean University of
China, Qingdao, Shandong, China
Received 3 August 2014; revised 29 September 2014; accepted 17 October 2014
KEYWORDS
Automated;
Data mining;
K-Nearest Neighbor;
On-line;
Real-Time
Abstract The major problem of many on-line web sites is the presentation of
many choices to the client at a time; this usually results to strenuous and time
consuming task in nding the right product or information on the site. In this
work, we present a study of automatic web usage data mining and recommendation system based on current user behavior through his/her click stream data on
the newly developed Really Simple Syndication (RSS) reader website, in order to
provide relevant information to the individual without explicitly asking for it.
The K-Nearest-Neighbor (KNN) classication method has been trained to be
used on-line and in Real-Time to identify clients/visitors click stream data,
matching it to a particular user group and recommend a tailored browsing
option that meet the need of the specic user at a particular time. To achieve this,
web users RSS address le was extracted, cleansed, formatted and grouped into
meaningful session and data mart was developed. Our result shows that the
K-Nearest Neighbor classier is transparent, consistent, straightforward, simple
* Corresponding author.
E-mail address: i@yangyongquan.com (Y. Yongquan).
Peer review under responsibility of King Saud University.
91
to understand, high tendency to possess desirable qualities and easy to implement than most other machine learning techniques specically when there is little
or no prior knowledge about data distribution.
2015 Production and hosting by Elsevier B.V. on behalf of King Saud University.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).
1. Introduction
Data mining is the extraction of knowledge from large amount of observational
data sets, to discover unsuspected relationship and pattern hidden in data, summarize the data in novel ways to make it understandable and useful to the data users
[13,31,2]. Web usage mining is the application of data mining technique to automatically discover and extract useful information from a particular web site
[2,22,30].
The term web mining was believed to have rst came to be in 1996 by Etzioni in
his paper titled The World Wide Web: Quagmire or Gold mine and since then
attention of researchers world over has been shifted to this important research
area [26]. In recent years, there has been an explosive growth in the number of
researches in the area of web mining, specically of web usage mining.
According to Federico and Pier [9], over 400 papers have been published on
web mining since the early paper published in 1990s.
The Really Simple Syndication (RSS) reader website was developed for the purpose of reading dailies news on-line across the Globe, but lack ways of identifying
client navigation pattern and cannot provide satisfactory Real-Time response to
the client needs, so, nding the appropriate news becomes time consuming which
makes the benet of on-line services to become limited. The study aimed at designing and developing an automatic, online, Real-Time web usage data mining and
recommendation system based on data mart technology. The system is able to
observe users/clients navigation behavior by acting upon the users click stream
data on the RSS reader web site, so as to recommend a unique set of objects that
satises the need of an active user in a Real-Time, online basis. The user access and
navigation pattern model are extracted from the historical access data recorded in
the users RSS address URL le, using appropriate data mining techniques.
The K-Nearest Neighbor classication method was used online and in RealTime to exploit web usage data mining technique to identify clients/visitors click
stream data matching it to a particular user group and recommend a tailored
browsing option that meet the need of the specic user at a given time [24]. For
instance, if a user seems to be searching for politics news on china daily on his/
her visit to the RSS reader site, more politics news headlines from other dailies
such as CNN politics news will be recommended to the user with the required feed
needed to be added to his/her prole in order to access such news headlines asides
his/her originally requested news. This is aimed at assisting the user to get relevant
92
information without explicitly asking for it, so as to ease and fasten navigation on
the site without too many choices being presented to the user at a time, More so,
the study will assist the web designer and administrator to re-arrange the content
of the web site in order to improve the impressiveness of the web site by providing
online Real-Time recommendation to the client.
To achieve this, the web users RSS address URL le was extracted, cleansed,
formatted and grouped into meaningful session for data mining analysis, data
mart was developed, this is as a result of the fact that the raw URL le extracted
is not well structured to be used directly for data mining [19]. In designing the
data mart, the process of URL data acquisition and model extraction was implemented using database management software specically the Structured Query
Language, MySQL 2008 [20]. The process of the development of the automatic
Real-Time web usage mining and recommendation application was done by
adopting the Java programming language with NetBeans as the editor and compiler [21]. The MATLAB software was used for interpretation and graphical presentation of the result obtained [17]. A thorough presentation of the
experimental result was done in order to assist the site designer and administrator to improve the content and impressiveness of the said RSS reader site. Fig. 1
showing the architecture of the overall system can be seen in Supplementary
material.
2. Related work
This section reviews some related works pertinent to this study, the review is
specically organized into subsections as follows:
2.1. Web data mining
Zdravko and Daniel [31], described web data mining as application of data mining
techniques to discover patterns in web content, structure and usage. It is a branch
of applied articial intelligence that deals with storage, retrieval and analysis of
web log les in order to discover users accessing and usage pattern of web pages
[26].
2.2. Forms of data mining system
Two forms of data mining tasks were identied by researchers over the years, these
includes; predictive and descriptive [13,1,7].
In predictive data mining task, inference is performed on current data in a database in order to predict future values of interest while in descriptive task, data in a
database are classied by characterizing the general properties of the data, it nds
pattern describing the data in the database so as to present the interpretation to
the user [13,1,8].
93
94
usage pattern discovery module, while in our work, user proles are determined
online, thereby making real time response and recommendation faster.
The path analysis model: Resul and Ibrahim [26], in their work used the path
analysis method to investigate the URL information of access to the Firat
University web server, web log le so as to discover user accessing pattern of
the web pages, in order to improve the impressiveness of the web site. They explain
further that, the application of path analysis method provides a count of number
of time a link occur in the data set, together with the list of association rules which
help to understand the path that users follow as they navigate through the Firat
University web site.
The Path analysis model is based on information from the clients previous navigation behavior, the method provides a count of number of time a link occur in
the dataset. Though our work shares the same goal of recommendation but again
differs in its approach, which is based on users current navigation behaviors
rather than previous navigation behavior as in path analysis method.
Bayesian classier model: Decision rule and Bayesian network, support vector
machine and classication tree techniques were used by Rivas et al. [27], to model
accidents and incidents in two companies in order to identify the cause of accident.
Data were collected through interview and modeled. The experimental result was
compared with statistics techniques, which shows that the Bayesian network and
the other methods applied are more superior than the statistics technique. Rivas
et al. [27], stated further that the Bayesian/K2 network is of advantage as it allows
what-if analysis on data, which make the data to be deeply explored.
In theory, Bayesian classier is said to have minimum error rate in comparison
with all other classier but in practice this is not always the case, due to inaccuracy
in assumptions made for its use, such as class conditional independency and the lack
of available probability data which is usually not the case when using KNN method.
The K-Nearest Neighbor (KNN): Many researchers have attempted to use KNearest Neighbor classier for pattern recognition and classication in which a
specic test tuple is compared with a set of training tuples that are similar to it.
[12], in their own work introduced the theory of fuzzy set into K-Nearest
Neighbor technique to develop a fuzzy version of the algorithm. The result of comparing the fuzzy version with the Crisp version shows that the fuzzy algorithm dominates its counterpart in terms of low error rate. In the work of [11]. The K-Nearest
Neighbor algorithm was used alongside with ve other classication methods to
combine mining of web server logs and web contents for classifying users navigation
pattern and predict users future request. The result shows that the KNN outperformed three of the other algorithms, while two of them performed uniformly. It
was also observed that KNN archives the highest F-Score and A(c) on the training
set among the six algorithms. [25], as well adopted the KNN classier to predict protein cellular localization site. The result of the test using stratied crossvalidation
shows the KNN classier to perform better than the other methods which includes
binary decision tree classier and the na ve Bayesian classiers.
95
2.4. Justication for using KNN algorithm over other existing algorithm
The K-Nearest Neighbor (K-NN) algorithm is one of the simplest methods for
solving classication problems; it often yields competitive results and has signicant advantages over several other data mining methods. Our work is therefore
based on the need to establish a exible, transparent, consistent straightforward,
simple to understand and easy to implement approach. This is achieved through
the application of K-Nearest Neighbor technique, which we have tested and
proved to be able to overcome some of the problems associated with other available algorithms. It is able to achieve these by the following:
Overcoming scalability problem common to many existing data mining methods such as decision tree technique, through its capability in handling training
data that are too large to t in memory.
The use of simple Euclidean distance to measure the similarities between training tuples and the test tuples in the absence of prior knowledge about distribution of data, therefore makes its implementation easy.
Reducing error rate caused by inaccuracy in assumptions made for usage of
other technique such as the Na ve Bayesian classication technique, such as
class conditional independency and the lack of available probability data which
is usually not the case when using KNN method.
Providing a faster and more accurate recommendation to the client with desirable qualities as a result of straightforward application of similarity or distance
for the purpose of classication.
2.5. Signicance of the study
Available published literature makes it clear that though web based recommendation systems are increasingly common, there still available many problem areas
calling for solutions. The fact is that most existing works lack scalability and capability when dealing with on-line, Real-Time search driven web sites, more so, the
recommendation quality and accuracy of some are doubtful, since they mostly
relied on historical information based on clients previous visit to the site, rather
than his immediate requirement. Some recommendation systems as well, create
a lot of bottleneck through system computing load when handling scaled web site
at peak visiting time thereby slowing down the recommendation process.
To solve the above issues the following solutions were made through our
system.
Scalability problems common to many existing recommendation system were
overcome through combine on-line pattern discovery and pattern matching
for real time recommendation, in this regard our algorithm works better than
decision tree algorithm.
96
Our result indicates that the adoption of the K-NN model can lead to a more
accurate recommendation that outperformed many other existing models. In
most cases the precision rate or quality of recommendation by our system is
equal to or better than 70%, meaning that over 70% of product recommended
to a client will be in line with his immediate requirement, making support to the
browsing process more genuine rather than a simple reminder of what the user
was interested in on his previous visit to the site as seen in path analysis
technique.
Our recommendation engine collects the active users click stream data, matches
it to a particular users group in order to generate a set of recommendation to
the client at a faster rate, therefore overcoming the problem of bottleneck
caused by system computing load when dealing with scaled web sites at a peak
visiting time, as it is in many existing data mining methods.
Our system provides a precise recommendation to the client based on his current navigation pattern, thereby overcoming time wastage in nding the right
product or information caused by presentation of many irrelevant choices to
the client at a time as it is in many existing systems.
Hence, the proposed approach is capable of addressing the issues and provides
a straightforward, simple to understand and easy to implement web usage classication and recommendation model.
3. Methodology
This section presents detail description of the realization and implementation of
web usage data mining system. The presentation of the application of the proposed methodology for the analysis of users RSS address le of the RSS reader
website was showcased. We have developed an online, Real-Time recommendation expert system that can assist the web designer and administrator to improve
the content, presentation and impressiveness of their website by recommending a
unique set of objects that satises the need of active user based on the users current click stream.
3.1. Overview of steps in performing web usage data mining task
Data mining task can be categorized into different stages based on the objective of
the individual analyzing the data [1,7].
The overview of the task for each steps is presented in detail in four subsections
as follows:
3.1.1. Data acquisition, preprocessing and data mart development
Data acquisition: This refers to the collection of data for mining purpose, and this
is usually the rst task in web mining application [6]. The said data can be
97
collected from three main source which includes (i) web server (ii) proxy server and
(iii) web client [9]. In this study, the web server source was chosen for the fact that
it is the richest and most common data source, more so, it can be used to collect
large amount of information from the log les and databases they represent. The
user prole information, the access and navigation pattern or model are extracted
from the historical access data recorded in the RSS reader site, users address database. The data are so voluminous as it contains so many detailed information such
as date, time in which activities occur, savers name, IP address, user name, password, dailies name, required feed, news headlines, and contents, as recorded in the
database le. In fact, the original document is about 5285 pages.
Data pre-processing: In the original database le extracted, not all the information are valid for web usage data mining, we only need entries that contain relevant information. The original le is usually made up of text les that contains
large volume of information concerning queries made to the web server in which
in most instance contains irrelevant, incomplete and misleading information for
mining purpose [30,11]. Resul and Ibrahim [26], described data preprocessing as
the cleansing, formatting and grouping of web log les into meaningful session
for the sole aim of utilizing it for web usage mining.
Data cleansing: Data cleansing is the stage in which irrelevant/noisy entries are
eliminated from the log le [18]. For this work the following operations were carried out: (i) Removal of entries with Error or Failure status. (ii) Removal of
requests executed by automated programs such as some access records that are
automatically generated by the search engine agent from access log le and proxies. (iii) Identication and removal of request for picture les associated with
request for a page and request include Java scripts (.js), and style sheet le (iv)
Removal of entries with unsuccessful HTTP status code, etc.
Data mart development: Two crown corporation [29], explained that data mart
is a logical subset of data warehouse. If the data warehouse DBMS can support
more resources, that will be required of the data mining operation, otherwise a
separate data mining database will be required. Since the raw log le is usually
not a good starting point for data mining operation, the development of a data
mart of log data is required for the data mining operation. In this work a separate
data mart of users RSS address URL was developed using relational database
Management software MySQL [20,19].
3.1.2. Transaction identication
There is need for a mechanism to distinguish different users so as to analyze users
access behavior [11]. Transaction identication is meant to create meaningful clusters of references for each user. Xuejuu et al. [30], stated that a user navigation
behavior can be represented as a series of click operations by the user in time
sequence, usually call click stream, which can further be divided into units of click
descriptions usually referred to as session or visit.
98
99
100
In general term, The Euclidean distance between two Tuples for instance
X1 = (x11, x12, . . ., x1n) and X2 = (x21, x22, . . ., x2n) will be,
q
Xn
distx1 ; x2
x1i x2i 2
3:2
i1
Eq. (3.2) is applicable to numeric attribute, in which we take the difference between
each corresponding values of attributes tuple x1 and x2, square the result and add
them all together then get the square root of the accumulated result this gives us the
distance between the two points x1 and x2 [13,14]. In order to prevent attributes
with initially large ranges from outweighing attributes with initial smaller ranges,
there is a need to normalize values of each attributes before applying Eq. (3.2).
The minmax normalization can be applied to transform for instance value V of
a numeric attribute A to V1 in the range [0,1] by using the expression
V minA
3:3
maxA minA
minA and maxA are attribute A, minimum and maximum values [13].
In K-NN, classication, all neighboring points that are nearest to the test tuple
are encapsulated and recommendation is made based on the closest distance to the
test tuple, this can be dened as follows:
Let C be the predicted class
V1
3:4
The nearest tuple is determined by the closest distance to the test tuple. The K-NN
rule is to assign to a test tuple the majority category label of its K-Nearest training
tuple [14].
3.3.1. Computing distance for categorical attribute
A categorical attribute is a nonnumeric attribute such as color and object name.
To calculate the distance, we simply compare the corresponding values of the attributes in tuple x1 with that of x2, if the values are the same, then the difference is
taken to be zero(0), otherwise the difference is taken to be one(1). For instance, if
two users, x1 and x2 click stream on the RSS reader site is both sport news category, then the difference is zero(0), but if tuple x1 is sport and tuple x2 is politics,
then the difference is taken to be one(1) [13].
3.3.2. Missing values
If the value of a given attribute A is missing in tuple x1 or x2 or both, for categorical value, if either or both values are missing we take the difference to be one(1),
in numeric attribute if x1 and x2 values are missing we also take the difference to
be one(1), if only one value is missing and the other is present and normalized we
can consider the difference to be |1 V1| or |0 V1| whichever is greater is chosen
[13].
101
102
Table 1
Users
Dailys name
News category
Class
X1
X2
X4
X5
X6
X7
X8
X9
X10
X11
X3
CNN news
China daily
CNN news
Punch ng
Thisday news
Vanguard news
Complete football
Vanguard news
China daily
Thisday news
Punch ng
World
Business
Politics
Entertainment
Politics
Sports
Sport
Politics
Politics
World
Politics
www.*world
www.*business
www.*politics
www.*entertainment
www.*politics
www.*sports
www.*sports
www.*politics
www.*politics
www.*world
www.*politics
World
Business
Politics
Entertainment
Politics
Sports
Sports
Politics
Politics
World
?
Remember, for a categorical attribute as in Table 1, the difference (x11, x31) can be
computed by simply compare the corresponding value of the attributes in tuple x1
with that of x3 as explained previously. If the values are the same then the difference is taken to be zero(0), otherwise, the difference is taken to be one(1). So, for
(x1,1 and x3,1) ie. (CNN news and Punch ng), the difference is 1, for (x12 and x32)
ie. (World and Politics) the difference is 1, likewise for (x13 and x33) ie., (www.*world and www.*politics) the difference is 1 as well, therefore,
q
2
2
2
distx1 ; x3 x1;1 x3;1 x1;2 x3;2 x1;3 x3;3 this gives :
distx1 ; x3
p2
p
1 12 12 3 1:73205081
Repeating the same process in our example for all other tuple x2, x4, . . ., x11, the
result of these calculation produced a stream of data as shown in Table 2, which
shows the users sorted by their Euclidean distance to the user x3 to be classied.
The 1-NN approach simply picks the user with minimum distance to x3,
(the rst one from the top of the list) and uses its class label politics to predict
the class of x3, therefore recommends similar news headlines of Politics as in
user x9 class.
Table 2
User
Class
Distance to user x3
x9
x10
x6
x4
x5
x2
x7
x8
x1
x11
Politics
Politics
Politics
Politics
Entertainment
Business
Sports
Sports
World
World
1.000000000
1.000000000
1.000000000
1.141421356
1.141421356
1.732050810
1.732050810
1.732050810
1.732050810
1.732050810
103
104
predictive ability of a model will be questionable and cannot be used for prediction, if the model performs well in the training set but performs poorly in the test
validation data set or new data set.
4.1. Presentation of result
Example 2: Using the data in Table 1, this time around assuming the class of user
X7 is unknown. We can determine the class of user x7 based on his current click
stream information by computing the Euclidean distance between the user x7
and all other users as we did in example 1
X1 = (CNN news, World, www.*world)
X7 = (Vanguard news, Sports, www.*sports)
q
Pn
2
dist(x1,x7) =
i3 x1i x7i
Being categorical attributes,
Differences (x1,1 x7,1) = 1, (x1,2 x7,2) = 1, (x1,3 x7,3) = 1
Therefore applying
Eq. (3.2) wep
have
p
dist(x1,x7) = 12 + 12 + 12 = 3 = 1.73205081
Repeating the whole process for all the available users produced a stream of
data as in Table 3.
4.2. Analysis of the results
The MATLAB code [17,23], that was used for graphical analysis of the experimental result from Tables 2 and 3 as shown in Fig. 4 and Figs. 5 and 6 is available on
request.
In order to model the users click stream in the RSS readers web site, The
K-Nearest Neighbor classication technique of data mining was applied on the
extracted users RSS address database. The data set was produced by computing
the Euclidean distance between the test tuple and the training tuples as shown
in example 1 and example 2 and data set presented in Tables 2 and 3 respectively.
Table 3
User
Class
Distance to user X7
x8
x9
x1
x2
x3
x4
x5
x6
x10
x11
Sports
Politics
World
Business
Politics
Politics
Entertainment
Politics
Politics
World
1.0000000000
1.1414213560
1.7320508100
1.7320508100
1.7320508100
1.7320508100
1.7320508100
1.7320508100
1.7320508100
1.7320508100
Figure 4
105
Graph showing Euclidean distance from the other User/Neighbor to user X3 and X7.
The K-Nearest Neighbor classier predicts the class label with class Ci for
which Ci = {x 2 Cp; d(x,xi) 6 d(x,xm), i #m} for the unknown user class ie., the
1-NN classication simply picks the user with minimum distance to users X3
and X7 as the case may be (ie. The rst user from the top of the list), in Table 2
for user X3 and Table 3 for user X7 respectively and use their class labels to predict
the class of X3 and X7 respectively, therefore, recommend similar news headline of
politics for user X3 as in user X9 class from Table 2 and Sports for user X7 as in
user X8 class from Table 3 as shown in Fig. 4, Figs. 5 and 6 respectively. Figs. 5
and 6 can be found in Supplementary material.
5. Summary of ndings
Different criteria can be used to determine the quality and efciency of a particular
web site, which includes the following: contents, presentation, ease of usage, ease
of accessing required information, waiting time of users, to mention just a few. In
this study a novel approach is presented to classify users based on their current
click stream, matching it to a particular user group popularly referred to as
Nearest neighbor and recommend a tailored browsing option that satises the
needs of the active user at a particular time, by applying the web usage data mining technique to extract knowledge required for providing Real-Time recommendation services on the web site.
We have conducted experiments on our designed experimental system. The data
set used in the system is the RSS user access database for a two months period,
which was extracted, pre-processed and grouped into meaningful sessions and
data mart was developed. The K-Nearest Neighbor classication technique was
used to investigate the URL information of the RSS users address database of
the RSS reader site as stored in the data mart created. Evaluating sample testing
106
session, the results are presented and analyzed. The results of our experiment indicate that the adoption of K-Nearest Neighbor model can lead to more accurate
recommendation that outperformed other classication algorithms. In most cases
the precision rate or quality of recommendation is equal to or better than 70%,
this means that over 70% of news recommended to a client will be in line with
his immediate requirement, making support to the browsing process more genuine, rather than a simple reminder of what the user was interested in on his previous visit to the site as seen in path analysis technique.
The ndings of the experimental study can now be used by the designer and
administrator of the web site to plan the upgrade and improvement of the web site,
in order to ease navigation on the site without too many choices at a time as well
as meeting their needed information without expecting them to ask for it explicitly,
therefore improving the impressiveness of the web site.
6. Conclusion
Our work provides a basis for automatic Real-Time recommendation system. The
system performs classication of users on the simulated active sessions extracted
from testing sessions by collecting active users click stream and matches this with
similar class in the data mart, so as to generate a set of recommendations to the
client in a Real-Time basis.
The result of our experiment shows that an automatic Real-Time recommendation engine powered by K-NN classication model implemented with Euclidean
distance method is capable of producing useful and a quite good and accurate
classications and recommendations to the client at any time based on his immediate requirement rather than information based on his previous visit to the site.
7. Recommendation for future work
Our designed system is a proof-of-concept, prototype of idea for using web
usage data mining with K-NN technique, and there are some aspects in which
it can be improved in any future work. The study could be taken much further
by investigating the users RSS address URL of the RSS reader in a continuous
basis.
More research also need to be carried out on many other data mining techniques, comparing the result with this model, so as to determine the most effective
model in handling a problem of this nature in the nearest future.
Acknowledgments
This research is supported by the Fundamental Research Funds for the Central
Universities (No. 201413065) and The National Key Technology R&D Program
(No. 2012BAH117F03)
107
108
[23] I.O. Ogbonaya, Introduction to Matlab/Simulink, for engineers and scientist, 2nd edition., John Jacobs
Classic Publishers Ltd, Enugu, Nigeria, 2008.
[24] C. Padraig, J.D. Sarah, K-Nearest Neighbor Classier. Technical Report UCD-CSI-2007-4, University
College Dublin, 2007.
[25] H. Paul, N. Kenta, Better Prediction of Protein Cellular Localization Sites with the K-Nearest Neighbor
Classier, ISMB-97, Proceeding of America Association for Articial Intelligence, USA, 1997, pp. 147152.
[26] D. Resul, T. Ibrahim, Creating meaningful data from web log for improving the impressiveness of a web site
by using path analysis method, Journal of expert system with applications 36 (2008) (2008) 66356644, http://
dx.doi.org/10.1016/j.eswa.2008.08.067.
[27] T. Rivas, M. Paz, J.E. Martins, J.M. Matias, J.F. Gracia, J. Taboadas, Explaining and predicting workplace
accidents using data-mining Techniques, Journal of Reliable Engineering and System safety 96 (7) (2011)
739747, http://dx.doi.org/10.1016/j.ress.2011.03.006.
[28] L. Shu-Hsien, C. Pei-Hui, H. Pei-Yuan, Data mining techniques and applications- A decade review from
2000 to 2011, Journal of expert system with applications 39 (2012) (2012) 1130311311, http://dx.doi.org/
10.1016/j.eswa.2012.02.063.
[29] Two Crown Corporation, Introduction to Data Mining and Knowledge Discovery, third ed., Two crown
corporation, 10500 Falls Road, Potamac, MD 20854, USA, 1999, pp. 540.
[30] Z. Xuejuu, E. John, H. Jenny, Personalised online sales using web usage data mining, J. Comput. Ind. 58
(2007) (2007) 772782, http://dx.doi.org/10.1016/j.compind.2007.02.004.
[31] M. Zdravko, T.L. Daniel, Data mining the Web, Uncovering patterns in Web content, structure, and usage,
John Wiley & sons Inc., New Jersey, USA, 2007, p. 115132.